Hands-on fundamentals of high-performance computing

This hands-on session introduces basic usage of the Snellius supercomputer at SURF. You will learn how to:

• connect to Snellius,
• understand the Linux environment,
• navigate the file system,
• compile a simple parallel program,
• submit and monitor jobs using the SLURM workload manager.

1. Accessing Snellius and exploring the Linux environment

Objective: Connect to Snellius and retrieve basic system and user information. Verify your access and become familiar with the Linux environment. Log this information in your home directory, in a file called <username>.dat.

#	Question	Command	Environment variable
1	What is your username?	whoami	echo $USER
2	What is your current working directory?	pwd	echo $PWD
3	What is your home directory?	none	echo $HOME
4	Which login node are you on?	hostname	echo $HOSTNAME
5	When did you last log in to Snellius?	last -n 1 $USER	none
6	How much storage quota do you have?	myquota	none
7	What is your compute budget and usage?	accinfo	none

Connecting to Snellius

To access Snellius, connect via SSH (troubleshooting guide):

ssh <username>@snellius.surf.nl

Creating your first file

After connecting to Snellius, your Shell will be located at your home directory: /home/<username>. This is a common setup in Linux environments. If you are using an scur login, your home directory should be empty. Let's create your first file, called .dat. There, we will log some of the environment values during the following steps.

[<username>@int6 ~]$ touch <username>.dat
[<username>@int6 ~]$ ls
<username>.dat

🖥️ Editing files in the Shell

There are multiple ways to edit files in a Shell. Two common options are:

• Shell editors like vim or nano, for manual editing.
• Output redirection using > for overwriting or >> for appending.

Exploring the Linux environment

When you log in to Snellius (or any Linux-based system), you interact with the system through a shell. The shell is a command-line interpreter that reads and executes commands. It also gives access to environment variables—named values used to configure your session or store user-specific details.

For example, you can check your username using the environment variable $USER. This will return your SURF username, which is set automatically at login.

[<username>@int6 ~]$ echo $USER
<username>

You will encounter environment variables often in this course, for tasks such as selecting compilers, accessing file paths, or configuring MPI jobs.

📦 Why do environment variables exist?

You’re working on a shared system (Snellius), with thousands of users and hundreds of programs. Every time you run a program, it needs to know who you are, where your files are, what software you use, what language or region settings you prefer.

Typing these details every single time would be painful and error-prone. That’s where environment variables come in: they act as a set of predefined, user-specific values that make the shell context-aware and help programs behave correctly without manual input every time.

Let’s say John Doe wants to create a folder for logs and save a file there. John could hardcode:

mkdir /home/johnd/logs
echo "My log entry" > /home/johnd/logs/run.log

But this will fail for everyone else, because /home/johnd/ only works for John. Instead, John could use the environment variable $HOME, which automatically expands to the correct home directory for any user:

mkdir -p $HOME/logs
echo "My log entry" > $HOME/logs/run.log

✅ Now the script works for everyone, not just John. This is essential in shared environments, job scripts, and automation.